]> git.vanrenterghem.biz Git - git.ikiwiki.info.git/log
git.ikiwiki.info.git
7 years agogit: change calling convention of safe_git to have named arguments
Simon McVittie [Wed, 28 Dec 2016 19:26:33 +0000 (19:26 +0000)]
git: change calling convention of safe_git to have named arguments

(cherry picked from commit 7f2235478d4331b8738e9a9dc8d6d6c08461cd5c)

7 years agogit: Do the revert operation in a secondary working tree
Simon McVittie [Wed, 28 Dec 2016 18:04:34 +0000 (18:04 +0000)]
git: Do the revert operation in a secondary working tree

This avoids leaving the git directory in an inconsistent state if the
host system is rebooted while we are processing a revert.

(cherry picked from commit 7e84a1f9d88a1f546188a28524fedbdf53bb8729)

7 years agogit: Turn $git_dir into a stack
Simon McVittie [Wed, 28 Dec 2016 18:02:59 +0000 (18:02 +0000)]
git: Turn $git_dir into a stack

This will be necessary when we use a secondary working tree to do
reverts without leaving the primary working tree in an inconsistent
state.

(cherry picked from commit 39b8931ad31fe6b48afdc570caa459a0996c2092)

7 years agoAdd automated test for using the CGI with git, including CVE-2016-10026
Simon McVittie [Mon, 26 Dec 2016 18:24:19 +0000 (18:24 +0000)]
Add automated test for using the CGI with git, including CVE-2016-10026

(cherry picked from commit fa64672d40f877f3bf9cf245cda0cc3f3837c50c)

7 years agoTry revert operations (on a branch) before approving them
Simon McVittie [Mon, 26 Dec 2016 18:45:02 +0000 (18:45 +0000)]
Try revert operations (on a branch) before approving them

Otherwise, we have a time-of-check/time-of-use vulnerability:
rcs_preprevert previously looked at what changed in the commit we are
reverting, not at what would result from reverting it now. In
particular, if some files were renamed since the commit we are
reverting, a revert of changes that were within the designated
subdirectory and allowed by check_canchange() might now affect
files that are outside the designated subdirectory or disallowed
by check_canchange().

OVE-20161226-0002

7 years agogit: do not fail to commit if committer is anonymous
Simon McVittie [Mon, 26 Dec 2016 18:21:37 +0000 (18:21 +0000)]
git: do not fail to commit if committer is anonymous

(cherry picked from commit c86046090e1dc31035e4db12e4f29562634d621e)

7 years agogit: don't issue a warning if rcsinfo is undefined
Simon McVittie [Mon, 26 Dec 2016 18:20:41 +0000 (18:20 +0000)]
git: don't issue a warning if rcsinfo is undefined

The intention here seems to be that $prev may be undefined, and the
only way that can legitimately happen is for $params{token} to be
undefined too.

(cherry picked from commit fd14cd2a4355684951bb40a1e72e8b0960a674fd)

7 years agoUse git log --no-renames for recentchanges
Simon McVittie [Sat, 3 Sep 2016 22:29:37 +0000 (23:29 +0100)]
Use git log --no-renames for recentchanges

Otherwise, recent git releases show renames as renames, and we do not
see that newdir/test5 was affected.

Bug-Debian: https://bugs.debian.org/835612
(cherry picked from commit 276f0cf57861418fae5b4db8446d3d1098130cdf)

7 years agoCorrectly handle filenames starting with a dash in add/rm/mv.
Florian Wagner [Thu, 17 Mar 2016 11:20:30 +0000 (12:20 +0100)]
Correctly handle filenames starting with a dash in add/rm/mv.

(cherry picked from commit bbdba8d770b73bc44f55219615b360484b7d240f)

7 years agoensure_committer: don't do anything if we have the environment variables
Simon McVittie [Mon, 30 Nov 2015 20:46:12 +0000 (20:46 +0000)]
ensure_committer: don't do anything if we have the environment variables

(cherry picked from commit 1f635c6dcaeff8f869f874f659da875c4e7f1863)

7 years agoDon't memoize ensure_committer
Simon McVittie [Mon, 30 Nov 2015 20:45:38 +0000 (20:45 +0000)]
Don't memoize ensure_committer

This makes it harder to test, and if we're invoking git anyway,
a couple of extra subprocesses are no big deal.

(cherry picked from commit 8550c397016bd66095f24de64b077526e08bbab2)

7 years agogit: if no committer identity is known, set it to "IkiWiki <ikiwiki.info>" in .git...
Simon McVittie [Mon, 26 Dec 2016 18:19:52 +0000 (18:19 +0000)]
git: if no committer identity is known, set it to "IkiWiki <ikiwiki.info>" in .git/config

This resolves commit errors in versions of git that require a non-trivial
committer identity.

(cherry picked from commit ed1e1ebe70c8aec06a759d8cd0168f242d28ac17)

7 years agoRevert "Tell `git revert` not to follow renames (CVE-2016-10026)"
Simon McVittie [Mon, 26 Dec 2016 18:18:45 +0000 (18:18 +0000)]
Revert "Tell `git revert` not to follow renames (CVE-2016-10026)"

This doesn't work prior to git 2.8: `git revert` silently ignores the
option and succeeds. We will have to fix CVE-2016-10026 some other way.

This reverts commit bb5cf4a0940b8fd2750c6175adb15382b84c71e2.

7 years agoTell `git revert` not to follow renames (CVE-2016-10026)
Simon McVittie [Mon, 19 Dec 2016 13:48:56 +0000 (13:48 +0000)]
Tell `git revert` not to follow renames (CVE-2016-10026)

Otherwise, we have an authorization bypass vulnerability: rcs_preprevert
looks at what changed in the commit we are reverting, not at what would
result from reverting it now. In particular, if some files were renamed
since the commit we are reverting, a revert of changes that were within
the designated subdirectory and allowed by check_canchange() might now
affect files that are outside the designated subdirectory or disallowed
by check_canchange().

Signed-off-by: Simon McVittie <smcv@debian.org>
7 years agoUpdate changelog for img backport
Simon McVittie [Wed, 11 Jan 2017 15:26:30 +0000 (15:26 +0000)]
Update changelog for img backport

7 years agoimg test: exercise upper-case extensions for image files
Simon McVittie [Mon, 9 May 2016 20:53:10 +0000 (21:53 +0100)]
img test: exercise upper-case extensions for image files

(cherry picked from commit 6879e3782854f336bb2aa192a36a049695953245)

7 years agoDetect image type from .JPG just like .jpg (etc.).
Amitai Schlair [Sun, 8 May 2016 22:26:15 +0000 (18:26 -0400)]
Detect image type from .JPG just like .jpg (etc.).

(cherry picked from commit 89af9ecc575c3e7223b0e6bea57f82872f5caa9f)

7 years agoUpdate changelog
Simon McVittie [Wed, 11 Jan 2017 14:57:55 +0000 (14:57 +0000)]
Update changelog

7 years agoCGI, attachment, passwordauth: harden against repeated parameters
Simon McVittie [Wed, 11 Jan 2017 13:22:03 +0000 (13:22 +0000)]
CGI, attachment, passwordauth: harden against repeated parameters

These instances of code similar to OVE-20170111-0001 are not believed
to be exploitable, because defined(), length(), setpassword(),
userinfo_set() and the binary "." operator all have prototypes that
force the relevant argument to be evaluated in scalar context. However,
using a safer idiom makes mistakes less likely.

7 years agoForce CGI::FormBuilder->field to scalar context where necessary
Simon McVittie [Sat, 24 Dec 2016 15:03:51 +0000 (15:03 +0000)]
Force CGI::FormBuilder->field to scalar context where necessary

CGI::FormBuilder->field has behaviour similar to the CGI.pm misfeature
we avoided in f4ec7b0. Force it into scalar context where it is used
in an argument list.

This prevents two (relatively minor) commit metadata forgery
vulnerabilities:

* In the comments plugin, an attacker who was able to post a comment
  could give it a user-specified author and author-URL even if the wiki
  configuration did not allow for that, by crafting multiple values
  to other fields.
* In the editpage plugin, an attacker who was able to edit a page
  could potentially forge commit authorship by crafting multiple values
  for the rcsinfo field.

The remaining plugins changed in this commit appear to have been
protected by use of explicit scalar prototypes for the called functions,
but have been changed anyway to make them more obviously correct.
In particular, checkpassword() in passwordauth has a known prototype,
so an attacker cannot trick it into treating multiple values of the
name field as being the username, password and field to check for.

OVE-20161226-0001

(cherry picked from commit c1120bbbe8fdea20cf64fa12247f4f4a4006c834)

7 years agopasswordauth: avoid userinfo forgery via repeated email parameter
Simon McVittie [Wed, 11 Jan 2017 13:19:13 +0000 (13:19 +0000)]
passwordauth: avoid userinfo forgery via repeated email parameter

OVE-20170111-0001

7 years agot/passwordauth.t: new automated test for passwordauth
Simon McVittie [Wed, 11 Jan 2017 13:16:37 +0000 (13:16 +0000)]
t/passwordauth.t: new automated test for passwordauth

In particular this includes an exploit for OVE-20170111-0001.

7 years agopasswordauth: prevent authentication bypass via multiple name parameters
Simon McVittie [Wed, 11 Jan 2017 13:12:50 +0000 (13:12 +0000)]
passwordauth: prevent authentication bypass via multiple name parameters

Calling CGI::FormBuilder::field with a name argument in list context
returns zero or more user-specified values of the named field, even
if that field was not declared as supporting multiple values.
Passing the result of field as a function parameter counts as list
context. This is the same bad behaviour that is now discouraged
for CGI::param.

In this case we pass the multiple values to CGI::Session::param.
That accessor has six possible calling conventions, of which four are
documented. If an attacker passes (2*n + 1) values for the 'name'
field, for example name=a&name=b&name=c, we end up in one of the
undocumented calling conventions for param:

    # equivalent to: (name => 'a', b => 'c')
    $session->param('name', 'a', 'b', 'c')

and the 'b' session parameter is unexpectedly set to an
attacker-specified value.

In particular, if an attacker "bob" specifies
name=bob&name=name&name=alice, then authentication is carried out
for "bob" but the CGI::Session ends up containing {name => 'alice'},
an authentication bypass vulnerability.

This vulnerability is tracked as OVE-20170111-0001.

7 years agoReference CVE-2016-4561 in 3.20141016.3 changelog
Simon McVittie [Mon, 9 May 2016 21:34:58 +0000 (22:34 +0100)]
Reference CVE-2016-4561 in 3.20141016.3 changelog

8 years ago3.20141016.3 (for jessie-security) debian/3.20141016.3
Simon McVittie [Fri, 6 May 2016 06:57:45 +0000 (07:57 +0100)]
3.20141016.3 (for jessie-security)

8 years agoDo not recommend mimetype(image/*)
Simon McVittie [Fri, 6 May 2016 06:46:58 +0000 (07:46 +0100)]
Do not recommend mimetype(image/*)

Not all image file types are safe for general use: in particular,
image/svg+xml is known to be vulnerable to CVE-2016-3714 under some
ImageMagick configurations.

8 years agoDocument the security fixes in this release
Simon McVittie [Fri, 6 May 2016 06:45:26 +0000 (07:45 +0100)]
Document the security fixes in this release

8 years agoadd more details of CVE-2015-2793
smcv [Tue, 14 Apr 2015 17:38:13 +0000 (13:38 -0400)]
add more details of CVE-2015-2793

8 years agoupdate for recent XSS
Joey Hess [Mon, 30 Mar 2015 15:31:59 +0000 (11:31 -0400)]
update for recent XSS

8 years agoimg: make img_allowed_formats case-insensitive
Simon McVittie [Fri, 6 May 2016 06:32:17 +0000 (07:32 +0100)]
img: make img_allowed_formats case-insensitive

8 years agoupdate test suite for svg passthrough by img directive
Joey Hess [Fri, 6 May 2016 00:44:11 +0000 (20:44 -0400)]
update test suite for svg passthrough by img directive

Remove build dependency libmagickcore-6.q16-2-extra which was only there
for this test.

8 years agoimg: Add back support for SVG images, bypassing ImageMagick and simply passing the...
Simon McVittie [Fri, 6 May 2016 05:57:12 +0000 (06:57 +0100)]
img: Add back support for SVG images, bypassing ImageMagick and simply passing the SVG through to the browser

SVG scaling by img directives has subtly changed; where before size=wxh
would preserve aspect ratio, this cannot be done when passing them through
and so specifying both a width and height can change the SVG's aspect
ratio.

(This patch looks significantly more complex than it was, because a large
block of code had to be indented.)

[smcv: drop trailing whitespace, fix some spelling]

8 years agoChangelog
Simon McVittie [Thu, 5 May 2016 22:45:16 +0000 (23:45 +0100)]
Changelog

8 years agoimg: check magic number before giving common formats to ImageMagick
Simon McVittie [Thu, 5 May 2016 22:17:45 +0000 (23:17 +0100)]
img: check magic number before giving common formats to ImageMagick

This mitigates CVE-2016-3714 and similar vulnerabilities by
avoiding passing obviously-wrong input to ImageMagick decoders.

8 years agoimg: restrict to JPEG, PNG and GIF images by default
Simon McVittie [Wed, 4 May 2016 07:54:19 +0000 (08:54 +0100)]
img: restrict to JPEG, PNG and GIF images by default

This mitigates CVE-2016-3714. Wiki administrators who know that they
have prevented arbitrary code execution via other formats can re-enable
the other formats if desired.

8 years agoUpdate structure of img test from master
Simon McVittie [Mon, 30 Nov 2015 17:33:00 +0000 (17:33 +0000)]
Update structure of img test from master

Originally from commit cdfb4ab "Run autopkgtest tests using autodep8 and
the pkg-perl team's infrastructure", cherry-picked here to be able to
apply subsequent test coverage extensions in this test.

8 years agoimg test: set old timestamp on source file that will change
Simon McVittie [Sun, 14 Jun 2015 17:13:17 +0000 (18:13 +0100)]
img test: set old timestamp on source file that will change

This is so that the test will pass even if it takes less than 1 second.

8 years agoimg test: skip testing PDFs if unsupported
Simon McVittie [Mon, 18 Jan 2016 09:19:42 +0000 (09:19 +0000)]
img test: skip testing PDFs if unsupported

8 years agoimg test: use the right filenames when testing that deletion occurs
Simon McVittie [Thu, 5 May 2016 22:24:17 +0000 (23:24 +0100)]
img test: use the right filenames when testing that deletion occurs

Also use a less misleading name for the sample SVG: it is no longer empty.
Since commit 105f285a it has contained a blue square.

8 years agoimg: force common Web formats to be interpreted according to extension
Simon McVittie [Wed, 4 May 2016 07:52:40 +0000 (08:52 +0100)]
img: force common Web formats to be interpreted according to extension

A site administrator might unwisely set allowed_attachments to
something like '*.jpg or *.png'; if they do, an attacker could attach,
for example, a SVG file named attachment.jpg.

This mitigates CVE-2016-3714.

8 years agoHTML-escape error messages (OVE-20160505-0012)
Simon McVittie [Wed, 4 May 2016 07:46:02 +0000 (08:46 +0100)]
HTML-escape error messages (OVE-20160505-0012)

The instance in cgierror() is a potential cross-site scripting attack,
because an attacker could conceivably cause some module to raise an
exception that includes attacker-supplied HTML in its message, for
example via a crafted filename. (OVE-20160505-0012)

The instances in preprocess() is just correctness. It is not a
cross-site scripting attack, because an attacker could equally well
write the desired HTML themselves; the sanitize hook is what
protects us from cross-site scripting here.

8 years agoimg: stop ImageMagick trying to be clever if filenames contain a colon
Simon McVittie [Sat, 13 Jun 2015 19:00:08 +0000 (20:00 +0100)]
img: stop ImageMagick trying to be clever if filenames contain a colon

$im->Read() takes a filename-like argument with several sets of special
syntax. Most of the possible metacharacters are escaped by the
default `wiki_file_chars` (and in any case not particularly disruptive),
but the colon ":" is not.

It seems the way to force ImageMagick to treat colons within the
filename as literal is to prepend a colon, so do that.

9 years agoRelease 3.20141016.2 debian/3.20141016.2
Simon McVittie [Sun, 29 Mar 2015 21:29:32 +0000 (22:29 +0100)]
Release 3.20141016.2

9 years agoFix XSS in openid selector. Thanks, Raghav Bisht.
Joey Hess [Fri, 27 Mar 2015 16:17:39 +0000 (12:17 -0400)]
Fix XSS in openid selector. Thanks, Raghav Bisht.

Conflicts:
debian/changelog
doc/bugs/XSS_Alert...__33____33____33__.html

9 years agonote that the two potential FTBFSs do not actually affect buildds debian/3.20141016.1
Simon McVittie [Wed, 7 Jan 2015 11:32:37 +0000 (11:32 +0000)]
note that the two potential FTBFSs do not actually affect buildds

The relevant tests are skipped on buildds because they don't have
libipc-run-perl or inkscape.

9 years agorelease candidate
Simon McVittie [Wed, 7 Jan 2015 11:10:32 +0000 (11:10 +0000)]
release candidate

9 years agoWork around Debian #771047: use a non-blank SVG for the regression test
Simon McVittie [Wed, 7 Jan 2015 11:08:31 +0000 (11:08 +0000)]
Work around Debian #771047: use a non-blank SVG for the regression test

Inkscape loses the bounding box of a SVG with no content when it
converts it to EPS, and ImageMagick does not have a special case for
converting SVG to PNG with Inkscape in one step (which Inkscape can do);
it prefers to convert SVG to EPS with Inkscape, then EPS to whatever.

9 years agoclose debian bug I opened about blogspam
Joey Hess [Fri, 2 Jan 2015 20:45:26 +0000 (16:45 -0400)]
close debian bug I opened about blogspam

9 years agoblogspam uses JSON instead of RPC::XML now.
Amitai Schlair [Sat, 3 Jan 2015 15:02:20 +0000 (10:02 -0500)]
blogspam uses JSON instead of RPC::XML now.

9 years agoUpdate blogspam to the 2.0 API.
Amitai Schlair [Fri, 2 Jan 2015 18:55:10 +0000 (13:55 -0500)]
Update blogspam to the 2.0 API.

9 years agoSet Debian package maintainer to Simon McVittie as I'm retiring from Debian.
Joey Hess [Sat, 8 Nov 2014 04:08:33 +0000 (00:08 -0400)]
Set Debian package maintainer to Simon McVittie as I'm retiring from Debian.

9 years agoAdd missing build-depends on libcgi-formbuilder-perl, needed for t/relativity.t
Joey Hess [Mon, 20 Oct 2014 16:28:54 +0000 (12:28 -0400)]
Add missing build-depends on libcgi-formbuilder-perl, needed for t/relativity.t

9 years agoMerge remote-tracking branch 'refs/remotes/dgit/dgit/sid' 3.20141016 debian/3.20141016
Simon McVittie [Thu, 16 Oct 2014 23:02:33 +0000 (00:02 +0100)]
Merge remote-tracking branch 'refs/remotes/dgit/dgit/sid'

9 years agorelease
Simon McVittie [Thu, 16 Oct 2014 22:28:35 +0000 (23:28 +0100)]
release

9 years agodebian: fix some wrong paths in the copyright file
Simon McVittie [Thu, 16 Oct 2014 22:28:23 +0000 (23:28 +0100)]
debian: fix some wrong paths in the copyright file

9 years agodebian: rename debian/link to debian/links so the intended symlinks appear
Simon McVittie [Thu, 16 Oct 2014 22:04:11 +0000 (23:04 +0100)]
debian: rename debian/link to debian/links so the intended symlinks appear

9 years agoclose a bug
Simon McVittie [Thu, 16 Oct 2014 22:03:48 +0000 (23:03 +0100)]
close a bug

9 years agoDrop unused python-support dependency
Simon McVittie [Thu, 16 Oct 2014 21:48:09 +0000 (22:48 +0100)]
Drop unused python-support dependency

9 years agochangelog so far
Simon McVittie [Thu, 16 Oct 2014 21:44:29 +0000 (22:44 +0100)]
changelog so far

9 years agobuild-depend on libcgi-pm-perl too, for tests
Simon McVittie [Thu, 16 Oct 2014 21:40:52 +0000 (22:40 +0100)]
build-depend on libcgi-pm-perl too, for tests

9 years agoExplicitly depend on CGI.pm, which is no longer in Perl core
Simon McVittie [Thu, 16 Oct 2014 08:45:36 +0000 (09:45 +0100)]
Explicitly depend on CGI.pm, which is no longer in Perl core

I was going to depend on the version that has CGI->param_fetch,
but that has been supported since 2.37, which is older than oldstable.

9 years agoIkiWiki::Plugin::openid: as a precaution, do not call non-coderefs
Amitai Schlair [Wed, 15 Oct 2014 22:52:43 +0000 (23:52 +0100)]
IkiWiki::Plugin::openid: as a precaution, do not call non-coderefs

We're running under "use strict" here, so if CGI->param's array-context
misbehaviour passes an extra non-ref parameter, it shouldn't be executed
anyway... but it's as well to be safe.

[commit message added by smcv]

9 years agoCall CGI->param_fetch instead of CGI->param in array context
Amitai Schlair [Wed, 15 Oct 2014 21:32:02 +0000 (22:32 +0100)]
Call CGI->param_fetch instead of CGI->param in array context

CGI->param has the misfeature that it is context-sensitive, and in
particular can expand to more than one scalar in function calls.
This led to a security vulnerability in Bugzilla, and recent versions
of CGI.pm will warn when it is used in this way.

In the situations where we do want to cope with more than one parameter
of the same name, CGI->param_fetch (which always returns an
array-reference) makes the intention clearer.

[commit message added by smcv]

9 years agoMake sure we do not pass multiple CGI parameters in function calls
Simon McVittie [Sat, 11 Oct 2014 08:28:22 +0000 (09:28 +0100)]
Make sure we do not pass multiple CGI parameters in function calls

When CGI->param is called in list context, such as in function
parameters, it expands to all the potentially multiple values
of the parameter: for instance, if we parse query string a=b&a=c&d=e
and call func($cgi->param('a')), that's equivalent to func('b', 'c').
Most of the functions we're calling do not expect that.

I do not believe this is an exploitable security vulnerability in
ikiwiki, but it was exploitable in Bugzilla.

9 years agoAdded a comment: It was an Apache problem...
https://www.google.com/accounts/o8/id?id=AItOawk8U772S3jDrZJCO0WA5WaDLjJv5mMl6Yw [Thu, 16 Oct 2014 14:57:26 +0000 (10:57 -0400)]
Added a comment: It was an Apache problem...

9 years agobranch
smcv [Thu, 16 Oct 2014 12:11:52 +0000 (08:11 -0400)]
branch

9 years agocomment
smcv [Thu, 16 Oct 2014 11:52:05 +0000 (07:52 -0400)]
comment

9 years agoReplace PayPal and Flattr buttons with text links
Simon McVittie [Wed, 15 Oct 2014 20:56:11 +0000 (21:56 +0100)]
Replace PayPal and Flattr buttons with text links

In particular, this avoids loading third-party resources from the
offline documentation (see
<https://lintian.debian.org/tags/privacy-breach-donation.html>).

9 years agomention pagespec_alias patches
http://anastigmatix.net/ [Thu, 16 Oct 2014 02:53:41 +0000 (22:53 -0400)]
mention pagespec_alias patches

9 years agoAdded a comment
smcv [Wed, 15 Oct 2014 23:30:22 +0000 (19:30 -0400)]
Added a comment

9 years agoAdded a comment
smcv [Wed, 15 Oct 2014 23:26:52 +0000 (19:26 -0400)]
Added a comment

9 years agoAdded a comment
openmedi [Wed, 15 Oct 2014 18:49:16 +0000 (14:49 -0400)]
Added a comment

9 years agoAdded a comment
https://www.google.com/accounts/o8/id?id=AItOawlcaGfdn9Kye1Gc8aGb67PDVQW4mKbQD7E [Wed, 15 Oct 2014 13:43:25 +0000 (09:43 -0400)]
Added a comment

9 years agoAdded a comment
openmedi [Wed, 15 Oct 2014 12:33:40 +0000 (08:33 -0400)]
Added a comment

9 years ago(no commit message)
https://www.google.com/accounts/o8/id?id=AItOawmbuZI4n1RsTe3Yeaqb5F-yhtR7a8BWEIE [Wed, 15 Oct 2014 04:18:10 +0000 (00:18 -0400)]

9 years agoas usual, macports hasn't moved
Amitai Schlair [Tue, 14 Oct 2014 22:46:41 +0000 (18:46 -0400)]
as usual, macports hasn't moved

9 years agoAdded a comment
https://www.google.com/accounts/o8/id?id=AItOawlcaGfdn9Kye1Gc8aGb67PDVQW4mKbQD7E [Tue, 14 Oct 2014 22:41:59 +0000 (18:41 -0400)]
Added a comment

9 years ago(no commit message)
https://www.google.com/accounts/o8/id?id=AItOawlobQ5j7hQVIGkwMWW3yKB_DWqthJcpnsQ [Tue, 14 Oct 2014 22:31:11 +0000 (18:31 -0400)]

9 years agoAdded a comment
https://www.google.com/accounts/o8/id?id=AItOawlcaGfdn9Kye1Gc8aGb67PDVQW4mKbQD7E [Tue, 14 Oct 2014 22:25:13 +0000 (18:25 -0400)]
Added a comment

9 years agoone report suffices; not yet clear there's a bug
Amitai Schlair [Tue, 14 Oct 2014 22:19:09 +0000 (18:19 -0400)]
one report suffices; not yet clear there's a bug

9 years ago(no commit message)
https://www.google.com/accounts/o8/id?id=AItOawk8U772S3jDrZJCO0WA5WaDLjJv5mMl6Yw [Tue, 14 Oct 2014 13:46:55 +0000 (09:46 -0400)]

9 years ago(no commit message)
https://www.google.com/accounts/o8/id?id=AItOawk8U772S3jDrZJCO0WA5WaDLjJv5mMl6Yw [Tue, 14 Oct 2014 13:20:24 +0000 (09:20 -0400)]

9 years agoclarify
Amitai Schlair [Mon, 13 Oct 2014 20:21:15 +0000 (16:21 -0400)]
clarify

9 years agofindings and questions
Amitai Schlair [Mon, 13 Oct 2014 20:13:11 +0000 (16:13 -0400)]
findings and questions

9 years agoDo not pass ignored sid parameter to checksessionexpiry
Simon McVittie [Sat, 11 Oct 2014 08:28:02 +0000 (09:28 +0100)]
Do not pass ignored sid parameter to checksessionexpiry

checksessionexpiry's signature changed from
(CGI::Session, CGI->param('sid')) to (CGI, CGI::Session) in commit
985b229b, but editpage still passed the sid as a useless third
parameter, and this was later cargo-culted into remove, rename and
recentchanges.

9 years agocomments: don't log remote IP address for signed-in users
Simon McVittie [Sun, 12 Oct 2014 17:03:28 +0000 (18:03 +0100)]
comments: don't log remote IP address for signed-in users

The intention was that signed-in users (for instance via httpauth,
passwordauth or openid) are already adequately identified, but
there's nothing to indicate who an anonymous commenter is unless
their IP address is recorded.

9 years agogoogle search plugin: use https for the search
Simon McVittie [Sun, 12 Oct 2014 16:57:14 +0000 (17:57 +0100)]
google search plugin: use https for the search

9 years agodefault User-Agent changed
smcv [Sun, 12 Oct 2014 16:49:24 +0000 (12:49 -0400)]
default User-Agent changed

9 years agoSet default User-Agent to something that doesn't mention libwww-perl
Simon McVittie [Sat, 11 Oct 2014 08:43:34 +0000 (09:43 +0100)]
Set default User-Agent to something that doesn't mention libwww-perl

It appears that both the open-source and proprietary rulesets for
ModSecurity default to blacklisting requests that say they are
from libwww-perl, presumably because some script kiddies use libwww-perl
and are too inept to set a User-Agent that is "too big to blacklist",
like Chrome or the iPhone browser or something. This seems doomed to
failure but whatever.

9 years agoremoved
smcv [Sun, 12 Oct 2014 16:43:14 +0000 (12:43 -0400)]
removed

9 years agoAdded a comment
smcv [Sun, 12 Oct 2014 16:42:54 +0000 (12:42 -0400)]
Added a comment

9 years agohelp Markdown make a list
Amitai Schlair [Sun, 12 Oct 2014 16:42:13 +0000 (12:42 -0400)]
help Markdown make a list

9 years agoAdded a comment: fixed in a recent release, I think
https://www.google.com/accounts/o8/id?id=AItOawlcaGfdn9Kye1Gc8aGb67PDVQW4mKbQD7E [Sun, 12 Oct 2014 16:40:18 +0000 (12:40 -0400)]
Added a comment: fixed in a recent release, I think

9 years ago(no commit message)
openmedi [Sun, 12 Oct 2014 16:06:59 +0000 (12:06 -0400)]

9 years agoReplace shebang paths with the build-time $(PERL).
Amitai Schlair [Sun, 12 Oct 2014 15:08:13 +0000 (11:08 -0400)]
Replace shebang paths with the build-time $(PERL).

On non-Debian systems, /usr/bin/perl might not be the best available
Perl interpreter. Use whichever perl was used to run Makefile.PL,
unless it was "/usr/bin/perl", in which case there's nothing to do.

9 years agoExtract test subs for each site. No change meant.
Amitai Schlair [Sun, 12 Oct 2014 13:30:31 +0000 (09:30 -0400)]
Extract test subs for each site. No change meant.

9 years agoExtract run_cgi(). No functional change intended.
Amitai Schlair [Sun, 12 Oct 2014 04:01:09 +0000 (00:01 -0400)]
Extract run_cgi(). No functional change intended.

9 years agoExtract check_generated_content(). Same output.
Amitai Schlair [Sat, 11 Oct 2014 13:52:21 +0000 (09:52 -0400)]
Extract check_generated_content(). Same output.

9 years agoExtract check_cgi_mode_bits(). No change intended.
Amitai Schlair [Sat, 11 Oct 2014 01:17:39 +0000 (21:17 -0400)]
Extract check_cgi_mode_bits(). No change intended.

9 years agoExtract thoroughly_rebuild(), a slight test change.
Amitai Schlair [Sat, 11 Oct 2014 00:40:24 +0000 (20:40 -0400)]
Extract thoroughly_rebuild(), a slight test change.

I didn't try to parameterize when a test should fail when we can't
remove ikiwiki.cgi because there already isn't one. (Hooray, natural
language.) Instead, we stop worrying about it and always tolerate
ENOENT.